# PINT：一种具有共享 Predictive Intent 的机器人操作统一模型

> 原文链接：https://robotics.hellobike.com/papers/2026-10-12-2610-16355
> 频道：papers | 更新日期：2026-10-12 | 标签：arxiv, cs.RO
> 核心摘要：PINT通过共享块级Predictive Intent协调语义推理、世界建模与动作生成，并在RoboTwin 2.0-Hard、LIBERO-Plus和真实双臂任务上验证。

# PINT：一种具有共享 Predictive Intent 的机器人操作统一模型

::: v-pre
> 原文链接：[PINT: A Unified Model with Shared Predictive Intent for Robot Manipulation](https://arxiv.org/abs/2610.16355)（arXiv 2610.16355 · 发布日期：2026-10-12）

**PINT: A Unified Model with Shared Predictive Intent for Robot Manipulation**

PINT通过共享块级Predictive Intent协调语义推理、世界建模与动作生成，并在RoboTwin 2.0-Hard、LIBERO-Plus和真实双臂任务上验证。

## 核心要点

- 现有多层共享注意力的统一融合方式，可能削弱VLM与VGM各自专长。
- PINT插入可学习Intent Tokens形成Predictive Intent，条件化VGM与动作专家。
- 两阶段训练先冻结VGM塑造意图，再固定意图并适配VGM和动作专家。
- 无具身预训练下，PINT提升OOD泛化，并报告多项基准和真实任务成功率。

**验证范围与局限**：验证范围为RoboTwin 2.0-Hard、LIBERO-Plus，以及五个真实世界双臂任务。

## 英文摘要 (Abstract)

Generalist robot manipulation can benefit from pretrained vision-language models (VLMs) for semantic reasoning and video generation models (VGMs) for world modeling. However, existing unified models that directly fuse them through shared attention at multiple layers can make it harder to draw on their specialized capabilities. We introduce PINT, a unified model that coordinates semantic reasoning, world modeling, and action generation through a shared chunk-level Predictive Intent while preserving expert-specific pathways. PINT inserts learnable Intent Tokens into the VLM sequence to form this representation, conditioning both the VGM and an Action Expert (AE). It replaces only the VGM's prompt conditioning and retains the VGM at inference, where the AE reads its representations through one-way attention. Rather than prescribing what Predictive Intent should encode, PINT learns it jointly from the native prediction objectives of the VLM, VGM, and AE, which provide complementary views of the same upcoming task execution. Training uses two stages: Stage 1 freezes the VGM so that future video prediction shapes Predictive Intent with task-relevant future dynamics, while Stage 2 fixes the learned Intent and adapts the VGM and AE. Without embodied pretraining, PINT achieves strong in-distribution (ID) performance and substantially improves out-of-distribution (OOD) generalization, reaching 15.5% on RoboTwin 2.0-Hard compared with 1.8% for Fast-WAM and 3.2% for StarVLA, and 79.5% on LIBERO-Plus compared with 51.5% and 75.0%, respectively. Across five real-world bimanual tasks, PINT reaches 68% ID and 24% OOD success, compared with 35%/9% for pretrained Motus and 67%/35% for pretrained $π_{0.5}$.

作者：Xinyu Guo, Chao Jing, Yi Li, Xingyu Chen, Zhengxing Wu
:::

## 原始参考来源 (Sources)

- [arXiv 2610.16355：PINT: A Unified Model with Shared Predictive Intent for Robot Manipulation](https://arxiv.org/abs/2610.16355)